2. Bootstrap for One Mean

Подписаться 24 тыс.

Просмотров 25 тыс.

50% 1

Bootstrapping uses the observed data to simulate resampling from the population. This produces a large number of bootstrap resamples. We can calculate a statistic for each bootstrap resample and use the distribution of the simulated statistics to approximate characteristics of the population. This bootstrapping process can help us construct a confidence interval for a population parameter, even when the population distribution is unknown. This video illustrates how to calculate a bootstrap confidence interval for a mean.

Опубликовано:

10 сен 2024

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист

Посмотреть позже

Комментарии : 9

@olivierderidder1928 6 лет назад

love your videos! I'm taking a data analyst nanodegree and some of the content was explained a bit superficial but your videos really helped close the gap!

@ProfessorKnudson 5 лет назад

Good to hear! Good luck with your nanodegree.

@LedCepelin 6 лет назад

Hey Professor Knudson, Just for my understanding. When we sample with replacement, that means in each bootstrapped mean, a particular baby(in this case) can be included multiple(and a random number of) times, and hence, the n=1009, but the means are different for each bootstrap. Am I getting this right?

@ProfessorKnudson 5 лет назад

Yes, the mean will (probably) be different from each iteration of the bootstrap. In a sample, some babies will be included more than once and others will not be included at all.

@SNPolka56 7 лет назад

Will the resample be less than 1009? As I understand as long as it n -1, then we are ok with replacement. If it is exactly 1009, then even with resampling we will have just one sample "i.e., the original sample". Professor Knudson, it would be great if you clarify this. Thanks

@ProfessorKnudson 7 лет назад

The idea of this is we are trying to simulate the sampling distribution for the mean when n=1009. Since the original sample size was 1009, each resample we take has size 1009. The resamples are slightly different from the original sample because we are sampling with replacement. Sampling with replacement means we record the measurement then allow ourselves to sample that one again. If your original sample had one measurement from Mahmoud, one measurement from Rita, etc then one of your resamples might have Mahmoud's measurement in there twice and Rita's in there zero times.

@JoyEdem 6 лет назад

Thank you so much, you're a great instructor!

@somcana 6 лет назад

when you say replacement, does that mean you are accessing the original data where where the sample was drawn from? or you are picking smaller samples within the 1009 sample?

@ProfessorKnudson 5 лет назад

With replacement: choose a baby, record the value, put the baby back. Repeat.